1990 LeSabre....dies if you take your foot off the accelerator

[Highway_101]

I have a new addition to my Buick fleet. It is a 1990 Le Sabre w/ 3800 engine. I will probably end up having to get the car scanned at a repair shop, but I thought I would look for suggestions first.

The car will start, but only if you depress the accelerator. The engine will only stay running if you keep your foot on the gas. Whether parked or going down the road, the car will die if your foot is off the accelerator. The check engine light is on. The previous owner drove the car like this for at least a year. Sometimes it will die and not restart for awhile.

Any suggestions before I haul it to the shop?

Many thanks,

[Reatta Man]

I would suspect the fuel pressure regulator, and/or the Idle Speed Control motor. Both are mounted on top of the engine; the regulator is on the fuel rail and the ISC motor is on the throttle body.

I'd recommend pulling the code or codes; that will give you a better idea. The computer on a '90 is not very sophisticated, and should at least get you started in the right direction.

While you are doing this, remove the accordian (rubber) tube that goes from the air filter box to the throttle body, and clean the throttle body with a good soaking of B-12 Carb cleaner with the engine off. That could also contribute to idling problems.

Good luck.

Joe

[Dave@Moon]

A code scanner for a pre-1996 car is very inexpensive (~$20), and it should point you in the right direction. I'd be willing to bet there are now a lot of codes to sort through if it was driven with the light on a year.

I'd be willing to be this is emission-related. Quite likely that a sensor was bad originally, which caused the engine to run rich. Driving it that way for a year probably lunched the catalytic converter, which is now plugged and not letting the engine idle. There may be other damage as well, i.e. I'd be surprised if the O2 sensor survived if this is what happened. Pulling the codes is your only recourse.

--Good Luck!

[NTX5467]

Main controller of idle speed is the Idle Air Control (also called Idle Speed Control). It's a stepper motor that runs the valve back and forth to control the air control through the idle air passages in the throttle body. This unit gets its commands from the Engine Control Module. Sometimes, they'll fail and hang up at mid travel. They should set a code too.

In that same relationship, the related passages in the throttle body should be open and clean, not just "clear". This is just as important as having a working valve.

The other issue might be a sticking or partially open EGR valve. You can remove the EGR valve, needing a new gasket to reinstall it, to see if the seating area of the pintle valve has accumulated deposits to prevent it from seating fully in the "off" mode. This can affect idle speed and quality too. Should also set a code too. If the EGR valve is not functioning, as in held just a hair open, it can cause an idle speed issue too.

The throttle body getting "coked" usually keeps the throttle valve from closing fully at idle. This can also interact with the other IAC and EGR issues too. In some cases, backflow of the EGR gasses as they might enter the air intake track, sometimes more upstream than you might suspect, can result in these deposits. In the early days, B-12 was an acceptable cleaning agent, but there are now more specialized Throttle Body Cleaners on the market which usually work better and quicker than B-12. In some respects, "something" is better than "nothing".

Probably just a little maintenance can take care of the problem. The EGR valve would be the most expensive item to replace, probably.

Enjoy!

NTX5467

[Highway_101]

Thanks for all the input guys. Here is what I have done so far....and by the way.....maybe I am mixing up my terms, but this is not a "throttle body" injected engine.....it is what I would call multi port, or sequential fuel injection. Also is a distributor-less (DIS) system. I have worked on many cars, but never one without a distributor.

Anyway, with that said..... I checked the fuel pressure with engine running. It measured 37 psi. I am told that the spec. is 40 to 47 psi, but I don't know that 3 psi low is that big of a deal. I then bypassed the regulator and hooked the gauge up to the end of the line. I expected a really high "shutoff" reading. I got only 10 psi. I think I messed up. I messed up even further because before realizing that that 10 psi reading is probably inaccurate, I went ahead and pulled the fuel tank to get the pump out. At least I didn't already install the new $135 pump. Next, I bought one of those cheap code readers. I am now trying to decide what to do next. Here are the codes:

12 --- Just tells that the diagnostics are working

14 --- Coolant temp sensor - signal voltage low

24 --- Vehicle speed sensor - open or short circuit problems

42 --- Electric spark timing (EST) circuit - open or shorted to ground during engine run. OR Direct ignition system (DIS) fault - bypass circuit open or shorted to ground during engine run OR Fuel cutoff relay circuit - open or shorted to ground.

I am guessing it is that last one (code 42) that is my main problem, but I am not sure the next logical step in diagnosis. At least this is a good learning experience.....I need to learn more about these newer ignition systems.

Any suggestions from this point? I guess I should put the fuel pump and tank back in so I can check things out like the EGR etc.

Thanks,

Michael

[Highway_101]

Okay, I'm stupid. I just wanted to clarify that I know what you guys are talking about when refering to the "throttle body" on a "sequential" or "multi-port" injected engine.

I haven't done anything since my last post. I am still processing all the information. I am hesitant to take further steps, because I know the cost of throwing parts at it can add up quick.

Thanks,

[Reatta Man]

Michael,

You're not stupid, just learning by asking.

Whether you are talking about a carb, multi-port or throttle body fuel injection system, they all have one large air opening, with a butterfly valve to control the flow of air. That is one of the parts that needs to be cleaned; specifically the butterfly valve itself (both sides) and the tunnel opening itself. With an Idle Speed Control motor, there is a smaller opening where air flows to be monitored, sending a signal to the computer.

On a lot of the 3.8s, there is a small plastic insert with one or two VERY thin wires over which the air flows. Those wires are fragile and should not be disturbed or handled roughly if the system is disassembled. However, the air opening can get dirty, and also needs to be cleaned.

When comparing a multi-port to a throttle body or TBI system, all you are talking about is the number of injectors that shoot a spray of fuel into the airstream. With a multi-port or MPI system, there is usually one injector per cylinder. Each injector is located as close as possible to the intake valve.

With a TBI system, there is usually one, sometimes two injectors that spray fuel into the airstream very similar to the way a carb sprays fuel through the jets. The difference is, any fuel injection system is more precise and efficient than any carb, and the fuel injection system operates under very high pressure. Also, the carb injects fuel based upon the vacuum created as air is drawn through the carb.

The cleaning that needs to be done with either system is inside the injector(s). Either due to age, mileage, poor fuel, trash or water in the fuel, the injectors become partially clogged, causing more of a stream of gas rather than a vapor. This stream doesn't ingnite very well, so the injector or injectors need to be cleaned with a strong solvent mixed with the fuel that dissolves the junk without dissolving the plastic in the injector. That is why you need a strong cleaner like BG 44K or Techron.

The final problem that is somewhat fuel-related is due to the fuel hitting a very hot intake valve. Often the first part of the fuel spray will partially dry as a carbon deposit, creating deposits. Those deposits interrupt the air flow, and often act like a sponge to absorb even more fuel. When your car's computer determines not enough fuel is being burned by the signal from the O2 sensor, it tells the injector to stay open longer and inject even more fuel. Gas mileage goes down, along with performance, because some of the fuel is being wasted on the dirty valves. Some valves can also not seat, or seal properly, which causes a small loss of compression, leading to even more power loss. If some of the hot gases from the compression stroke leaks back into the intake manifold, it gets everything even dirtier, much faster.

Cleaning the opening, intake valves, ISC motor, injectors and all the surfaces inside the intake system usually puts the engine back to factory specs. The exception would be if the contamination destroyed the O2 sensor or the catalytic converter. More money, and bigger problems.

Hope this helps.

Joe

[NTX5467]

As stated, the best way to expand your knowledge on these issues is to do the research yourself and then the fixes, provided you have the expertise and tools to do them.

On any fuel injected engine, fuel pressure is highly important. On the later model GM engines, if they do not have at least the min spec fuel pressure, the injectors will not "fire". A restricted fuel filter can contribute to these things too. I've seen some late model light duty trucks that had "extended start time", meaning you had to crank them for longer than normal before they started. With the fuel pressure gauge hooked up, instead of the fuel pressure building quickly, it'd start low and then gradually rise during the cranking time until when it hit 55psi, then at 56psi it would finally start.

Some later model Buicks have an access panel in the front of the trunk floor, so you don't have to remove the fuel tank to get to the fuel pump/sending unit. Not sure if yours does or does not have that access panel.

On many of the Mass Air Flow sensors, the wire(s) that string from side to side are actually how the computer measures air flow. The cooling effect of the inflowing air is countered by the computer to keep them at a particular temperature level. More voltage to keep them heated, more air flow. Seems like it's called a "bridge circuit"? If those wires get caked or coated, it can effect the readings. To counteract that, many of those cars (TPI Corvettes have it) have a "burnoff relay/module" to self-clean the wires each time the vehicle is started and probably before it goes into closed loop operation. In some cases, removing the screen in front of the wires (a "popular hot up mod") can expose them to conditions they were not meant to endure and/or cause the fuel curve map to shift from where it should be.

As with carbureted engines, fuel issues and ignition issues can produce similar situations. Knowing how the error codes can intereact or be caused by "side issues" can be necessary, but with a little common sense and a knowledge of how things work, they can be figured out. It still takes the correct amount of fuel and sparks at the right time to make things work "as designed".

One other neat diagnostic tool can be one of the infrared temperature "guns". For example, if the area the temp sensor lives is really up to temp, but the sensor tells the computer otherwise, then the sensor could be faulty. Usually, there is a coolant sensor for the computer and then another one that runs the gauge in the instrument cluster. Sometimes, its the same sensor, many times they are different and screw into different parts of the engine (i.e, thermostat housing or the cylinder head). The basic RayTech gun is about $75.00 or less and goes to 500 degrees F, but if you want the 1000 degree F readings, they'll cost a good bit more (the higher temps can be used to read exhaust manifold temperatures at each port, for example, which can be used to find cylinders with weak spark or otherwise).

ALSO be aware that in some of the earlier years, the trouble codes usually changed each year. Therefore, a particular code for one year's vehicle would mean something entirely different for the next year's vehicle. So make sure the code list you have is for the year of vehicle you actually have.

Don't get frustrated and enjoy the learning experience. Using the trouble codes is a good way to do things, other than chunking parts at it until you might "get lucky". Making sure the computer gets accurate information makes for a better running vehicle.

Enjoy!

NTX5467

[Highway_101]

Thanks so much guys for taking the time to post all that information. I do have the tools. I actually worked as an auto mechanic for about 3 years. I am a little embarrased to admit that since I obvioiusly wasn't a very good one. Diagnostics were never my strong point. I was more of a "tell me what to install boss" kind of mechanic. I have been meaning to purchase one of those ray tech temp guns. I also need to pick up a good multi-meter.

Anyway, here is an update on what I have found. One tower on the coil pack was very corroded. I have cleaned it, but need to repair the spark plug wire end. I pulled all the plugs and performed a compression test. They ranged from 132 to 170 psi. Kind of a broad range I think, but at least none are real low. The plugs look like they have been in for a while. The gap was as much as .010 too wide (spec is .060). Some of the plugs weren't real tight. Also, they smelled of fuel.

Regarding fuel pressure, I was told by a friend who is a shop owner that the "engine running" spec is 40-47 psi. My Buick was running at 37 psi. Should this be a concern, or is it too close to spec to worry about. Could even be a gage calibration problem I suppose.

I still have the fuel tank out and a new fuel pump still in the box. I don't want to waste $160 (for the pump) if I don't have to......but I hate to reinstall the original one and find out later that it is bad.

Well, I have to get ready for work (I work nights). I check for any responses after midnight.

Thanks again,

[Reatta Man]

On a 1990 car, if you have the tank already out, definitely replace the pump.

I also strongly recommend you replace the filter. A clogged filter will make the pump work much harder.

The wide compression range is something to be concerned about. If you are not loosing oil (more than about one quart every 1500 miles or so) then I would suspect the valves are not seating well on the cylinders with the lowest readings. You will definitely need to run a good cleaner through the fuel system to get those deposits off the valves.

Joe